The Rothman Index Does Not Predict a Successful Extubation in the Neurosurgical Critical Care Unit.

Background Identification of risk factors associated with successful extubation in neurosurgical critical care units (NSICUs) has been elusive due to the complex nature of neurocritical care injuries and patient factors. Traditional risk factors for extubation were shown to have poor predictive value in neurocritical care patients as compared to mixed ICU patients. The aim of this study was to determine if any risk factors, including the Rothman Index, could reliably predict successful extubation in a large sample size of neurocritical care patients. Methods We retrospectively analyzed 610 consecutively intubated patients in an NSICU while collecting variables of interest in airway management. Furthermore, Rothman Indices were collected immediately after intubations and extubations. A paired t-test of the immediate changes in Rothman Indices after airway manipulation was compared in patients who needed reintubation. In a smaller cohort of 88 patients, in whom complete data points existed for airway management, we performed a principal component analysis (PCA) to determine which risk factors were associated with extubation success when indexed with the magnitude of the Rothman Index. Results In 610 consecutively intubated patients, the mean pre-intubation Rothman Index average was 41.0 compared to the mean post-extubation Rothman Index was 35.4 (p<0.0001). Compared to those who were re-intubated, the Rothman Index did not correlate well with the prediction of extubation (p=0.355). Furthermore, an analysis of the PCA plot showed that a higher respiratory rate, longer length of stay, shorter length of intubation, and smaller cuff leak percentage were identified as risk factors associated with reintubation. Age and change in rapid shallow breathing index (RSBI) did not correlate with reintubation. Conclusion The Rothman Index does not predict extubation success in patients in an NSICU. Risk factors associated with reintubation were respiratory rate, length of stay, length of intubation, and cuff leak percentage. Reintubation rates in our single-center NSICU are on par with general critical care populations.